Link theory in manifolds (Q1369890): Difference between revisions

The book under review is based on the author's ``Habilitation''. ``Invariants of links in 3-manifolds have been defined and studied classically through algebraic topology. In the last decade ideas from singularity theory, quantum field theory and statistical mechanics gave rise to the new theories of Vassiliev- and quantum invariants. But the resulting combinatorial context and the topology of links are not nicely combined. The Conway polynomial \(\nabla_K(z)\) of links \(K\) in \(S^3\) is an invariant, which is well understood from both the classical and the modern viewpoint. It is combinatorially characterized by \(\nabla_{unknot} = 1\) and the Conway relation \(\nabla_{K_+} -\nabla_{K_-} =z\nabla_{K_0}\).'' ``In the early seventies John Conway suggested to generalize the Conway polynomial for the case of oriented 3-manifolds \(M\) in the following way: Consider the quotient of the free \(Z[z]\)-module on the set of isotopy classes of oriented links in \(M\) by the submodule generated by all elements \(K_+-K_- -zK_0\). Obviously the resulting Conway skein module is universal with respect to maps from the set of isotopy classes of oriented links in \(M\) to \(Z[z]\)-modules satisfying the Conway relation. The Conway skein module of \(S^3\) is free of rank 1 and generated by unknot (with the image of a link in the module given by its Conway polynomial multiplied with the generator). Conway skein modules of 3-dimensional handlebodies have been computed in [\textit{V. G. Turaev}, J. Sov. Math. 52, No. 1, 2799-2805 (1990); translation from Zap. Nauch. Semin. Leningr. Otd. Mat. Inst. Steklova 167, 79-89 (1988; Zbl 0673.57004)] and [\textit{J. H. Przytycki}, Topology '90, Contrib. Res. Semester Low Dimensional Topol., Columbus/OH (USA) 1990, Ohio State Univ. Math. Res. Inst. Publ. 1, 315-342 (1992; Zbl 0770.57005)] using diagrammatic tools. But there is no topological understanding for the resulting invariants like in \(S^3\). Conway skein modules are typical examples of combinatorial universal constructions using link theory in \(M\). But in order to achieve explicit results concerning link theory, one needs to compute the modules and to understand the resulting link invariants topologically. Our approach to link theory is motivated by the problems above.'' ``The plan of these notes is as follows: We will start with a general discussion of link map bordism in manifolds and prove a classification result in a range of dimensions. This is mainly to show that the basis of our discussion is not specifically 3-dimensional. Generalizations of many concepts to high dimensional link theory are evident. In chapter 2 we discuss bordism relations in 3-manifolds and provide the classification of oriented resp. framed and embedded resp. singular bordism of \(r\)-component 1-dimensional links in compact connected oriented 3-manifolds. Chapter 3 contains the fundamental concepts of the theory including the computation of the universal rings of Betti-trivial 3-manifolds. In chapter 4 we set up the framework for invariants of oriented surfaces and determine the relation with link isotopy. Chapter 5 shows how, following the classical line of argument, Seifert pairing defined by ring \(lk\)-maps give rise to polynomial invariants of links in 3-manifolds. In chapter 6, the characteristics of surfaces and links in 3-manifolds are defined in general. Then the discussion is specialized to Betti-trivial manifolds. We prove theorem 5 and discuss applications including the generalized band sum problem. Finally in chapter 7 we study the problem of Betti-trivial submanifolds of arbitrary connected compact oriented 3-manifolds. A process of cutting along oriented surfaces until the complement becomes Betti-trivial is considered. This gives rise to natural complexity invariants of oriented 3-manifolds, which are not Betti-trivial. The appendix contains definitions and results on inner homology and inner Betti-numbers, important computational tools in the study of manifolds with boundary. Moreover, we have included a comprehensive treatment of the homology of link exteriors in 3-manifolds. General proofs of these results, which are indispensable for the understanding of the theory, do not seem available in the literature''. To have a flavor of the book we describe below Theorem 5 which generalizes a result of \textit{M. Scharlemann} and \textit{A. Thompson} [Comment. Math. Helv. 64, No. 4, 527-535 (1989; Zbl 0693.57004)]. A compact, connected, oriented 3-manifold is called Betti-trivial if \(2b_1(M)=b_1(\partial M)\), where \(b_1\) is the first Betti-number (equivalently \(M\) can be embedded in a rational homology sphere or \(M\) has no embedded oriented non-separating closed surface). Theorem 5. Suppose \(M\) is irreducible Betti-trivial and with boundary components toral or two-spheres. Let \(K_b,K\) be the links, defined by twisting and cutting a link in a ball in \(M\). Then either \(\chi (K_b)\) does not depend on the twist and is \(\leq \chi (K) -1\), or there is a unique twist \(c\) with \(\chi (K_c) >\chi (K_b) =\chi (K) -1\) for all twists \(b\neq c\). The link characteristic \(\chi (K)\) of oriented links \(K\) is defined to be the maximal Euler characteristic of oriented surfaces without closed components bounding \(K\). Reviewer's remarks: Universal linking number analyzed by the author is closely related to the second skein module described in detail by the reviewer [A \(q\)-analogue of the first homology group of a 3-manifold, Contemp. Math. 214, 135-144 (1998)]. In particular, a compact, connected, oriented 3-manifold is Betti-trivial iff its second skein module is free.